MedSiLa captures the symptom, then reasons on the drug, in one pipeline. Built for oncology, where polypharmacy risk is highest and the margin for error is thinnest.
Oncology patients are among the most medically complex populations in healthcare. Multiple concurrent therapies, narrow therapeutic windows, and drug interactions with unpredictable timing mean adverse drug reactions are often caught late, after a patient has already been harmed or a treatment has already failed.
Most existing tools solve one piece of this: a symptom tracker that captures data but does not reason about drugs, or a drug interaction checker that flags a theoretical risk with no patient context. Neither closes the loop between what a patient is experiencing and what is causing it.
MedSiLa is a four-stage pipeline that turns patient-reported symptoms and clinical data into causality-adjudicated, actionable alerts.
Structured electronic patient-reported outcomes (ePRO) capture symptoms as they happen, not weeks later at the next visit.
Symptom data is enriched with the patient's full medication record through FHIR and SMART on FHIR integration, pulling in concurrent therapies, dosing, and timing directly from the EHR.
A causality-gated fusion engine evaluates each symptom against known drug-drug interaction pathways (CYP450, QTc), polypharmacy burden (Drug Burden Index), and adjudicated causality using established clinical frameworks (WHO-UMC, Naranjo, Liverpool, DIPS). Similar historical cases are retrieved to support the assessment.
Alerts are tiered and gated: immediate alert, hold-and-re-evaluate, or suppress-and-log, so care teams see what matters when it matters, without alert fatigue. Every output is informational, designed to support clinical judgment rather than replace it.
Oncology is not a market MedSiLa expanded into. It is the starting point.
MedSiLa's reasoning layer is grounded in the same frameworks clinicians already trust:
The core mechanism, a causality-gated fusion of interaction signals, polypharmacy burden, and patient context, is the subject of a pending patent application.
Oncology care teams and investors — we'd like to hear from you.